Sensor State Estimation for Accurate Machine Abnormality Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional systems face challenges in precisely determining the actual state of a work machine, as they often assume sensor values are correct without distinguishing between sensor abnormalities and machine abnormalities, leading to potential misinterpretation of the machine's operational state.

Innovation Solution

A state estimation device and method that estimates the state of sensors based on sensor values, distinguishing between sensor and machine abnormalities by analyzing the percentage of abnormal operations and using machine learning to generate reference data for accurate assessments.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If sensor values are used directly to determine machine state without verification, then the system operation is simple and fast, but the accuracy of abnormality detection deteriorates due to inability to distinguish sensor abnormalities from machine abnormalities

Engineering Contradiction:
Improveabnormality detection accuracyVSAvoidstate estimation system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces a state estimation device as an intermediary component between the sensor and the machine state determination system. This device estimates the sensor state by analyzing the relationship between sensor values and machine operation states, thereby mediating the information flow and enabling more accurate abnormality detection without requiring direct modification of the core machine control system.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements feedback by continuously monitoring sensor values, comparing them with estimated machine states, and using this information to adjust the state estimation. The state estimation device receives feedback from both sensor measurements and machine state observations, refining its estimation of sensor health status over time to improve detection accuracy.

Inventive Principle:
Principle #23Feedback

2Reliability

If conventional systems assume sensor values are correct, then the system complexity is low, but the reliability of state determination deteriorates due to potential sensor abnormalities being misinterpreted as machine abnormalities

Engineering Contradiction:
Improvestate determination reliabilityVSAvoidsensor state estimation complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The state estimation device performs self-service by autonomously estimating sensor state and detecting abnormalities without requiring external intervention or complex additional hardware. The system uses available data from sensor readings and machine state observations to self-determine sensor health status, thereby improving reliability through intelligent processing rather than through added physical complexity.

Inventive Principle:
Principle #25Self-service

3Loss of information

If sensor abnormalities are not distinguished from machine abnormalities, then the diagnostic process is simple, but the loss of information increases due to misinterpretation of the actual problem source

Engineering Contradiction:
Improveabnormality source informationVSAvoiddiagnostic operation simplicity
Core Design Contradiction:
Loss of informationVSEase of operation

Solution Approach 1:

The patent applies segmentation by separating the diagnostic function into distinct components: sensor state estimation and machine state determination. This segmentation allows the system to independently analyze whether an abnormality originates from the sensor or the machine, thereby preserving information about the abnormality source while maintaining operational simplicity through automated analysis.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3862829B1State estimation device, system, and manufacturing method
Publication Date: 2023.03.29 YASKAWA DENKI KK
  • EP3862829B1 patent drawingFigure 1
  • EP3862829B1 patent drawingFigure 2
  • EP3862829B1 patent drawingFigure 3

AI summary

A state estimation device is provided, which includes: a machine controlling unit for controlling a work machine based on a sensor value acquired from a sensor configured to output the sensor value related to an operation by the work machine; and a state estimation unit for estimating a state of the sensor based on the sensor value. In addition, a system is provided, which includes: the state estimation device; the work machine; and the sensor. In addition, a method of manufacturing a manufacture item by a work machine is provided, which includes: controlling the work machine based on a sensor value acquired from a sensor configured to output a sensor value related to an operation by the work machine on the manufacture item; and estimating a state of the sensor based on the sensor value.